crevasse splay processes and deposits in an ancient ...eprints.whiterose.ac.uk/117911/7/crevasse...
TRANSCRIPT
This is a repository copy of Crevasse splay processes and deposits in an ancient distributive fluvial system: The lower Beaufort Group, South Africa.
White Rose Research Online URL for this paper:http://eprints.whiterose.ac.uk/117911/
Version: Accepted Version
Article:
Gulliford, AR, Flint, SS and Hodgson, DM orcid.org/0000-0003-3711-635X (2017) Crevasse splay processes and deposits in an ancient distributive fluvial system: The lower Beaufort Group, South Africa. Sedimentary Geology, 358. pp. 1-18. ISSN 0037-0738
https://doi.org/10.1016/j.sedgeo.2017.06.005
© 2017 Elsevier B.V. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/
[email protected]://eprints.whiterose.ac.uk/
Reuse
Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of the full text version. This is indicated by the licence information on the White Rose Research Online record for the item.
Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [email protected] including the URL of the record and the reason for the withdrawal request.
1
Crevasse splay processes and deposits in an ancient distributive fluvial system: the lower 1
Beaufort Group, South Africa 2
3
Alice R. Gulliford1ゆゅ, Stephen S. Flint1, David M. Hodgson2 4
1Stratigraphy Group, School of Earth and Environmental Sciences, University of Manchester, 5
Oxford Road, Manchester M13 9PL, U.K. 6
2Stratigraphy Group, School of Earth and Environment, University of Leeds, Leeds LS2 9JT, 7
U.K. 8
ΏShell International Ltd., 40 Bank Street, Canary Wharf, London. E14 5NR, U.K. 9
10
e-mail: [email protected] 11
*Corresponding author. 12
2
Abstract 13
Up to 12% of the mud-prone, ephemeral distributive fluvial system stratigraphy in the 14
Permo-Triassic lower Beaufort Group, South Africa, comprises tabular fine-grained 15
sandstone to coarse-grained siltstone bodies, which are interpreted as proximal to distal 16
crevasse splay deposits. Crevasse splay sandstones predominantly exhibit ripple to climbing 17
ripple cross-lamination, with some structureless and planar laminated beds. A hierarchical 18
architectural scheme is adopted, in which each ~1 m thick crevasse splay elements extend 19
for tens to several hundreds of meters laterally, and stack with other splay elements to form 20
crevasse splay sets up to 4 m thick and several kilometers in width and length. Paleosols and 21
nodular horizons developed during periods, or in areas of reduced overbank flooding and 22
are used to subdivide the stratigraphy, separating crevasse splay sets. Deposits from 23
crevasse splays differ from frontal splays as their proximal deposits are much thinner and 24
narrower, with paleocurrents oblique to the main paleochannel. In order for crevasse splay 25
sets to develop, the parent channel belt and the location where crevasse splays form must 26
stay relatively fixed during a period of multiple flood events. Beaufort Group splays have 27
similar geometries to those of perennial rivers but exhibit more lateral variability in facies, 28
which is interpreted to be the result of more extreme fluctuations in discharge regime. 29
Sharp-based crevasse splay packages are associated with channel avulsion, but most are 30
characterized by a gradual coarsening upward, interpreted to represent progradation. The 31
dominance of progradational splays beneath channel belt deposits may be more 32
characteristic of progradational stratigraphy in a distributive fluvial system rather than 33
dominated by avulsion processes in a trunk river system. This stratigraphic motif may 34
3
therefore be an additional criterion for recognition of distributive fluvial systems in the 35
ancient record. 36
Keywords: crevasse splay; overbank; avulsion; distributive fluvial system; Karoo Basin 37
4
1. Introduction 38
Fluvial sedimentological and stratigraphic research tends to focus on the internal 39
architecture, geometry and stacking patterns of channel sandstone bodies, rather than their 40
adjacent overbank successions (e.g., Allen, 1983; Blakey and Gubitosa, 1984; Bridge and Tye, 41
2000; Gouw and Berendsen, 2007; Pranter et al., 2009; Jenson and Pedersen, 2010). This is 42
because channel-fills are coarser grained and better exposed, and channel elements are of 43
greater importance as hydrocarbon reservoirs than their finer grained, usually less 44
permeable, overbank counterparts. However, in aggradation-dominated systems, floodplain 45
and crevasse splay deposits form a key constituent of the overall stratigraphic succession 46
(Bristow et al., 1999). Local floodplain accommodation is controlled by the elevation of the 47
bank-full channel relative to its surrounding floodbasin (Wright and Marriott, 1993). River 48
avulsion processes, floodplain morphology, and controls on floodplain evolution have been 49
a focus for previous research on crevasse splay deposits (e.g., Tyler and Ethridge, 1983; 50
Nanson and Croke, 1992; Kraus and Aslan, 1993; Singh et al., 1993; Walling and He, 1998; 51
Kraus and Wells, 1999; Slingerland and Smith, 2004; Jones and Hajek, 2007; Abels et al., 52
2013; Hajek and Edmonds, 2014; Burns et al., 2017). Crevasse splay successions provide 53
evidence for the mechanisms of channel avulsion. For example, several authors have 54
identified distinct types of splay stratigraphy (Kraus and Wells, 1999; Mohrig et al., 2000; 55
Slingerland and Smith, 2004; Jones and Hajek, 2007) interpreted to represent either abrupt 56
or gradual to failed avulsion. 57
The majority of depositional models for crevasse splays are derived from ancient and 58
modern temperate to boreal climates where organic-rich strata dominate (e.g., Horne et al., 59
1978; Ethridge et al., 1981; Flores, 1983; Smith et al., 1989; Jørgensen and Fielding, 1996; 60
5
Bristow et al., 1999; Jerrett et al., 2011; Burns et al., 2017). Natural variability in river 61
avulsion processes, and the architecture and dimensions of crevasse splay deposits, are not 62
adequately constrained by these case studies (Nanson and Croke, 1992). Studies that focus 63
on overbank successions within other climatic regimes include O'Brien and Wells (1986) on 64
ephemeral stream crevasse splay deposits from the subtropical Clarence River of Australia, 65
and ephemeral terminal splay deposits from Lake Eyre, Australia (Payenberg et al., 2004; 66
Fisher et al., 2008). 67
In addition, the role of crevasse splay deposits in the distributive fluvial system (DFS) 68
paradigm (i.e., Nichols and Fisher, 2007; Hartley et al., 2010a, 2010b; Sambrook Smith et al., 69
2010; Weissmann et al., 2010; Fielding et al., 2012; Gulliford et al., 2014) has not been 70
examined specifically. From an applied perspective, research into the role of overbank 71
successions in hydrocarbon reservoir development has been limited (e.g., Mjøs et al., 1993; 72
Ambrose et al., 2008; Stuart et al., 2014; van Toorenenburg et al., 2016). However, recent 73
interest in fine-grained overbank successions has increased due to a growth in coal bed 74
methane extraction and the hydraulic fracturing of fluvial tight gas reservoirs (e.g., Pashin, 75
1998; Ayers, 2002; Shanley et al., 2004; Iwere et al., 2006). 76
The aim of this paper is to address the need for a more detailed analysis of 77
sedimentary facies, geometries, stacking patterns and evolutionary development of ancient 78
crevasse splay deposits from non-coal bearing strata. The large-scale outcrops from the 79
Permo-Triassic lower Beaufort Group of the SW Karoo Basin, South Africa are used to 80
address the following objectives: i) to propose a hierarchical architectural classification 81
scheme enabling the characterization of different overbank elements over a range of spatial 82
scales; ii) to test models that characterize splays in terms of either random overbank 83
flooding or attempted channel avulsion; iii) to draw comparisons between modern and 84
6
ancient crevasse splay datasets; and iv) to address the role of crevasse splay deposits in 85
distributive fluvial system models. 86
2. Geological setting 87
The Permo-Triassic Karoo Basin fill of South Africa (88
89 Fig. 1) contains deep-water submarine fans though submarine slope and shelf edge delta 90
deposits of the 2 km thick Ecca Group (Flint et al., 2011), overlain by the 5 km+ thick fluvial 91
7
Beaufort Group (Johnson et al., 1997) (92
93 Fig. 2). Subsidence at this time was generated by dynamic topography associated with 94
mantle flow processes (Pysklywec and Mitrovica, 1999; Tankard et al., 2009). 95
8
The lower 900 m of the Beaufort Group (Abrahamskraal Formation) forms the focus 96
of this study, and is exposed over an area of 900 km² (97
98
Fig. 1). Fluvial channel belt and overbank deposits are best observed around the edge of the 99
Great Escarpment and in river cliffs, road cuts and hill sides. Tectonic dip is 1-2° to the east. 100
9
The stratigraphy is floodplain dominated, with crevasse splay deposits forming 12% (101
102
Fig. 3). Previous studies have interpreted the lower Beaufort Group as an ephemeral fluvial 103
system (Smith, 1990a, 1990b, 1993a, 1993b). The abundance of planar and low angle 104
laminated lower fine- to medium-grained sandstone with parting lineation in the channel 105
belt deposits of the lower Beaufort Group has been interpreted to indicate upper flow 106
regime conditions in grain size-limited rivers (Turner, 1981; Fielding et al., 2009; Gulliford et 107
al., 2014; Wilson et al., 2014). The characteristics of splay deposits have been reported by 108
Stear (1980, 1983), Jordaan (1990) and Smith (1993a), and recent studies have interpreted 109
the lower Beaufort Group as a distributive fluvial system (Wilson et al., 2014; Gulliford et al., 110
2014). 111
10
3. Dataset and methods 112
The outcrop dataset comprises 33 sedimentary logs measured at centimeter-scale, high 113
resolution photo panel interpretations and maps of key surfaces supported by paleocurrent 114
measurements, from five main study sites (Supplementary Table 1). Apparent widths were 115
corrected to true widths using paleotransport directions relative to the outcrop orientation. 116
At the escarpment edge, sedimentary log sections through a ~330 m thick stratigraphic 117
interval from Verlatenkloof and Müller Canyon (118
119
11
Fig. 1; Supplementary Figs. 1, 2) at a scale of 1:50 capture sedimentary facies and 120
architectural elements (121
122
Fig. 3; Supplementary Spreadsheet 1). Grainsize was quantified in the field using a grainsize 123
chart, with siltstones determined according to competency, degree of fissile character and 124
weathering style, and claystones feeling smooth when rubbed against the teeth. 125
Adjacent to the town of Sutherland, three studies have been carried out to 126
document overbank facies relationships, geometries, and stacking patterns. The main 127
detailed study area at Sutherland East includes 14 sedimentary logs that were measured at 128
a scale of 1:25. Individual beds (labeled A to Z, up to ZD) were walked out between logged 129
sections and correlated on a photo panel. Sedimentary log thicknesses and paleocurrent 130
readings were geo-referenced in GoogleEarthTM to constrain spatial changes in architecture 131
and splay thickness. Additional data were collected from study sites at Klipkraal Farm and 132
12
Bﾉﾗﾏげゲ F;ヴﾏ ふ133
134
Fig. 1). 135
4. Facies 136
Within the Lower Abrahamskraal Formation, 11 overbank facies have been observed, 137
including six mudstone facies and five sandstone facies. The mudstone facies comprise 138
13
fissile purple mudstone, poorly-sorted purple siltstone, bright green massive mudstone, 139
poorly-sorted green-gray-blue siltstone, laminated organic-rich dark gray mudstone and 140
thinly-bedded coarse-grained siltstone and very fine-grained sandstone (141
142
Fig. 4A-F) (Gulliford et al., 2014, Table 2). The sandstone facies comprise ripple cross-143
laminated very fine- to fine-grained sandstone, structureless and normally graded very fine- 144
to fine-grained sandstone, planar laminated very fine- to fine-grained sandstone, low angle 145
(< 10°) cross-stratified very fine- to fine-grained sandstone and trough cross-stratified fine-146
grained sandstone. Relative proportions of each overbank-related mudstone and sandstone 147
14
facies averaged across all the study areas are provided in Table 1, and Figure 4 shows key 148
features. 149
5. Architectural Elements 150
The facies have been grouped into eight architectural elements based on schemes by Allen 151
(1983), Friend (1983), Miall (1985, 1988, 1996) and Colombera et al. (2012, 2013). These 152
schemes combine sedimentary facies associations with grain size and geometries. Overbank 153
15
architectural elements comprise crevasse splay deposits, floodplain fines and floodplain lake 154
deposits ( 155
16
Fig. 5). The proportion of fluvial-overbank architectural elements measured from 156
sedimentary logs at the Great Escarpment is presented in Figure 3, and each element is 157
described below. 158
5.1 Floodplain fines architectural element 159
Description: The floodplain fines architectural element includes fissile purple mudstone, 160
poorly sorted purple siltstone, bright green massive mudstone and poorly sorted green-161
gray-blue siltstone facies (162
163
17
Fig. 4A-D), as described in Table 2 of Gulliford et al. (2014). Typically, the facies are 164
moderately to highly bioturbated by roots and burrows (Bioturbation Index BI 3 to 4; Taylor 165
and Goldring, 1993), thus destroying internal laminae. Polygonal features (1 cm across) are 166
infilled by siltstone. Successions are up to 40 m thick with sharp bases and are overlain by 167
sandstones with either gradational or sharp contact. The green siltstone deposits (minor Fbg 168
[169
170
18
Fig. 4C]: Munsell distinction 5G3/6; Fggb [171
172
19
Fig. 4D]: Munsell distinction 10G5/2 to 5B5/2) are coarser grained than the purple siltstone 173
(Fp; [174
175
20
Fig. 4A]: Munsell distinction 5P1/6 to 10P1/6; Fpu [176
177
Fig. 4B]: Munsell distinction 5P2/2 to 10P2/2). 178
Interpretation: The root traces with reduction haloes and slickensides suggest prolonged 179
subaerial exposure following suspension settling on subaqueous floodplains (Gulliford et al., 180
2014). The infilled polygonal features are interpreted as desiccation cracks, consistent with 181
subaerial exposure. 182
21
The localized preservation of carbonate nodules and absence of a thick calcrete layer 183
supports an interpretation of moderate maturity paleosols (sensu Leeder, 1975; Nichols, 184
2009), and a periodic and localized net moisture deficit (Wright et al., 2000). The purple 185
coloration is indicative of hematite (Kraus and Hasiotis, 2006). Purple to green color changes 186
and mottling have been previously associated with alternating oxidizing and reducing 187
environments, signifying fluctuations in the height of the water table and variable fluvial 188
discharge (Stear, 1980; Dubiel, 1987; Wilson et al., 2014). The grain size decrease associated 189
with the color change from green to purple siltstone may indicate alternating overbank and 190
distal crevasse deposits (Kraus and Aslan, 1993; Kraus and Wells, 1999; Abels et al., 2013). 191
The wide range of floodplain facies is thought to reflect the complicated discharge 192
pattern of ephemeral rivers with standing water present for long periods in lakes and 193
shorter periods on the draining floodplain. 194
5.2 Floodplain lake architectural element 195
Description: The floodplain lake architectural element comprises only laminated organic-rich 196
dark gray facies of claystone and sometimes normally graded siltstone (Figs. 4E, 5; Table 1). 197
Bedsets are 0.5 m to several meters thick (typically < 1 m thick) and stack into successions 198
that extend over hundreds of meters, with a sharp basal contact and a gradational to sharp 199
top surface. Bioturbation is moderate to intense (BI 3 to 5; Taylor and Goldring, 1993), 200
including horizontal and vertical burrows. 201
Interpretation: Floodplain lake elements are interpreted to form through the settling from 202
suspension of fines on waterlogged floodplains in topographic lows or in lakes, associated 203
with poor drainage and high groundwater table. 204
22
5.3 Crevasse splay architectural element 205
Description: The crevasse splay architectural element includes sharp-based thinly-bedded 206
coarse-grained siltstone and very fine-grained sandstone deposits < 2 m thick, comprising < 207
5 individual normally graded beds (Figs. 4F, 5). Packages are tabular and extend laterally 208
over hundreds of meters to approximately three kilometers with sharp tops and low-relief 209
(cm-scale) erosional basal contacts (Jordaan, 1990), which are overlain by rare sub-angular 210
23
mudstone clasts (< 5 mm in diameter). Overall, packages typically coarsen-upward (211
212
Fig. 5). The main sedimentary structures include ripple to climbing-ripple cross-lamination, 213
with some planar lamination and structureless deposits. Bioturbation is sparse to intense (BI 214
24
1 to 5; Taylor and Goldring, 1993), with burrows, trackways (vertebrate and invertebrate) 215
and trails common (Smith, 1990b, 1993a; Gulliford, 2014; Wilson et al., 2014). Burrow traces 216
are widespread, consisting of both vertical and horizontal pipe burrows. Horizontal burrows 217
are commonly straight, sinuous or branched, and may be seen cross-cutting one another. 218
Vertical burrows are complex and variable, with either U-shaped, sub-vertical, bulbous, or 219
chambered morphologies (Gulliford, 2014). Scoyenia traces are common and typically found 220
intersecting one another, on the desiccated tops of splay deposits. 221
25
Lenticular sand bodies are 1 m to < 3 m thick (222
223
26
Fig. 5), single-storey and ~5-20 m wide with concave-up, scoured bases and sharp tops, and 224
wings in cross-section, as previously noted by Stear (1983). The lenticular bodies comprise 225
ripple cross-laminated sandstone (Sr), planar-laminated sandstone (Sh), structureless 226
sandstone (Sm), and low angle cross-stratified sandstone (Sl), with rare trough cross-227
stratified sandstone (St). 228
Interpretation: The laterally extensive packages dominated by tractional sedimentary 229
structures are interpreted as the products of rapid deposition from unconfined flow, and 230
therefore as crevasse splays. The presence of structureless sandstone and/or climbing ripple 231
cross-lamination is interpreted to indicate high rates of sediment fallout and tractional 232
deposition that is attributed to rapid expansion and deposition from moderate to low 233
concentration unconfined flows (Allen, 1973). The type of climbing ripple cross-lamination 234
produced depends on the sediment fallout rate from suspension (Jopling and Walker, 1968; 235
Allen, 1973). 236
The lenticular bodies are interpreted as crevasse channels. Crevasse splays and 237
crevasse channel-fills are located adjacent to larger channel belt sandstones (3-12 m thick). 238
These channel belts are coarser grained (i.e., upper fine-grained sandstone compared with 239
lower fine-grained sandstone) and they typically exhibit a complex internal architecture with 240
upper-flow-regime structures throughout. These characteristics are indicative of deposition 241
under highly variable flow conditions (Turner, 1981; Fielding, 2006; Fielding et al., 2009; 242
Gulliford et al., 2014; Wilson et al., 2014). Paleocurrent measurements obtained from ripple 243
and climbing ripple cross-lamination in crevasse splay sandstones (Table 2) are typically 244
oblique to orientations of channel belts, as reported by Gulliford et al. (2014). 245
27
5.4 Crevasse splay architecture 246
High-resolution 2-D analysis of two splay successions, from Klipkraal Farm (Figs. 6, 7) and 247
Bﾉﾗﾏげゲ F;ヴﾏ ふFigs. 8, 9), combine sedimentary facies observations (248
249
Fig. 4) with thickness and paleocurrent data (Table 2). The Klipkraal crevasse splay sand-250
body was analyzed in a section sub-ヮ;ヴ;ﾉﾉWﾉ デﾗ ヮ;ﾉWﾗaﾉﾗ┘ ┘ｴｷﾉW デｴW Bﾉﾗﾏげゲ W┝;ﾏヮﾉW ｷゲ 251
exposed perpendicular to paleoflow. Cross-sections illustrate horizontal and vertical 252
changes in sedimentary facies and thickness. In both cases, the deposits are dominated by 253
fine-grained sandstone with subordinate amounts of coarse-grained siltstone. 254
28
At Klipkraal Farm (255
256
Fig. 5), the crevasse splay sand body has a measured outcrop width of 100 m, but the 257
paleocurrent-corrected true width is 78 m (Table 2). A laterally extensive paleosol, 258
characterized by calcrete nodules, lies < 1 m above the top of the sandstone unit and was 259
29
used as a marker bed (260
261
Fig. 7). The sand body varies in thickness between 0.2 m and ~1.5 m, and has a weakly 262
erosional base into floodplain mudstone. The constituent very fine-grained sandstone beds 263
are structureless to normally graded, or show ripple and climbing-ripple cross-lamination, 264
with observed paleoflow ranging between 347-042° (n = 11) (265
266
30
Fig. 7). The NW margin of the splay is not exposed, but the body thins and fines 267
laterally SE of log KK-30 (268
269
Fig. 7). In the thicker, axial section, there is a coarsening- then fining-upward profile 270
(KK-27 and KK-29). Thin (millimeter-scale) claystone-siltstone laminae cap the splay and are 271
characterized by horizontal and vertical burrows. 272
Aデ Bﾉﾗﾏげゲ F;ヴﾏが a splay sandstone is characterized by climbing ripple cross-lamination 273
overlain by current ripple cross-lamination (Figs. 8, 9). Measured paleocurrents range 274
between 356-066° (n = 14), which indicates that the 190 m wide outcrop is perpendicular to 275
31
paleoflow. The body thins laterally from 2 m to 1 m (276
277
Fig. 9) and has a corrected width of 62 m (Table 2). Upper surfaces of the constituent 278
sandstone beds are rippled or flat, with intense bioturbation destroying internal 279
sedimentary structures. No apparent coarsening or fining trends are observed in the 280
floodplain mudstone beneath the splay. 281
The geometry, sedimentology and paleoenvironmental context of these two 282
sandstone units (Figs. 6-9) support an interpretation of crevasse splay deposits. They are 283
finer grained than the fills of surrounding channel belts, which range from very fine- to 284
lower medium-grained sandstone (Gulliford et al., 2014). The upward coarsening to fining 285
trend in the splay axis is interpreted to reflect the waxing to waning of flood energy during 286
deposition. The lateral decrease in grain size, both across strike and down dip, is attributed 287
to reducing energy of the floodwater as the flow expanded abruptly away from the crevasse 288
32
channel. The fine-grained drapes on top of crevasse splay deposits are interpreted as paleo-289
surfaces indicating periods of non-deposition (Stear, 1983; Smith, 1993a). 290
On a north-south oriented hillside at Sutherland East, sedimentary logs were 291
measured through a continuously exposed 4 m thick package of crevasse splay deposits 292
(Figs. 10, 11). A widespread nodular paleosol horizon forms a basal marker and logs were 293
measured up through the package to the erosional base of an overlying channel belt 294
sandstone. Six crevasse splay deposits with paleocurrent measurements from ripple and 295
climbing-ripple cross-laminated very fine-grained sandstone that range between 010-079° (n 296
= 12), are interpreted as genetically-related crevasse splays that extend for more than 700 297
m laterally (298
299
Fig. 11). The true extent of the package of crevasse splay deposits cannot be 300
determined due to exposure limitations. Individual splay deposits comprise numerous 301
lenticular sheets (< 0.2 m thick) of structureless or ripple cross-laminated siltstone and 302
sandstone, and are separated by 10 cm thick but laterally extensive floodplain mudstone 303
beds. The thin beds and abundance of coarse siltstone relative to (rare) fine-grained 304
sandstone are interpreted to represent a distal crevasse splay setting. 305
33
6. Discussion 306
6.1 Crevasse splay hierarchical scheme 307
A hierarchical classification of fluvial channel elements is widely employed (Miall, 1985, 308
1996; Payenberg et al., 2011; Ford and Pyles, 2014; Gulliford et al., 2014). The same 309
approach is more challenging to implement for overbank deposits because elements are 310
laterally extensive and fine-grained, there is an absence of prominent erosional bounding 311
surfaces, and their stratal relationship to parent channels is rarely preserved at outcrop. 312
However, a crevasse splay deposit hierarchical scheme is proposed here, from bed-scale 313
through crevasse splay element to crevasse splay set (314
315
Fig. 12) that uses the presence of paleosols to subdivide the stratigraphy. In aggrading 316
systems, deposits associated with crevasse and avulsion processes are commonly preserved 317
(Slingerland and Smith, 2004). The floodplain aggrades due to the growth and subsequent 318
abandonment of splay deposits (Smith et al., 1989), during periods when flow is diverted 319
from the main fluvial channel to the floodplain, or through a full channel avulsion. Channel 320
belt abandonment or avulsion to a position far away is represented by a paleosol or nodular 321
horizon in the overbank deposits (Willis and Behrensmeyer, 1994) where sediment-laden 322
34
flood waters rarely encroach. Therefore, the succession between two paleosols can be 323
considered to represent the increment of floodplain aggradation associated with a single 324
channel belt that is close enough to actively supply sediment to the floodplain. Assuming 325
that the rate of tectonic or compactional subsidence does not vary locally, then the 326
thickness of the floodplain succession between paleosols will decrease with distance from 327
the parent channel belt. 328
A single sandbody bounded by paleosols in a floodplain succession is a crevasse splay 329
element, and is deposited during an overbank flooding event whereby a levee is breached or 330
overtopped and material from within the channel builds up on the floodplain. The examples 331
of well-defined sand-bodies from Klipkraal (Figs. 6, 7) ;ﾐS Bﾉﾗﾏげゲ F;ヴﾏ ふFigs. 8, 9) are 332
interpreted as individual proximal crevasse splay elements bounded by paleosols. Stacked 333
crevasse splay elements with similar paleocurrents and scale, and with no significant 334
intercalated paleosols are defined as a genetically-related crevasse splay set that comprise 335
two or more crevasse splay elements (336
337
Fig. 12). Laterally, a crevasse splay set may thin and be expressed as a single crevasse splay 338
element bounded by paleosols. The Sutherland East dataset (Figs. 10, 11) provides an 339
example of a crevasse splay set that has significant vertical facies variability (Fig. 10B-G), 340
35
marked by repeated fining- and coarsening-upward profiles. These suggest either abrupt 341
initiation/termination of splays or complicated, probably compensational, stacking of splay 342
elements. Upward-coarsening above a slightly erosional base is interpreted as signifying 343
preliminary floodplain incision and splay growth, followed by fining-upwards as flood 344
conditions wane (Bridge, 1984). 345
The crevasse splay set at Sutherland East (Figs. 10, 11) is defined by a nodular horizon 346
below and paleosol above. Each crevasse splay element within the crevasse splay set is 347
much thinner (< 0.2 m thick) than crevasse splay elements observed at Klipkraal (Figs. 6, 7) 348
;ﾐS Bﾉﾗﾏげゲ F;ヴﾏ ふFｷｪゲく Βが Γ) with average grain size between coarse siltstone and very fine 349
sandstone. These characteristics together with abrupt vertical facies changes, associated 350
with stacked, heterolithic deposits, are interpreted as a distal crevasse splay set (351
352
Fig. 12). 353
36
Lower in the Klipkraal Farm and Great Escarpment stratigraphy (354
355
Fig. 1), paleosols are spaced at ~1.5 m intervals (0.5-2 m observed range, n = 12), with no 356
intercalated sand-prone splays. Based on the overbank hierarchical scheme outlined above, 357
these closely stacked paleosols are likely to have developed several kilometers from the 358
channel belt, away from active flooding. In the younger escarpment stratigraphy, multiple 359
crevasse splay elements are present between paleosols > 3 m apart; this architecture 360
indicates greater sediment supply to the floodplain, suggesting a position closer to the 361
parent channel. With reference to Kraus (1987) and Kraus and Aslan (1993), one or more 362
37
splays in a section between two paleosols should be time-equivalent to the deposition of a 363
single- or multi-storey channel belt, and by implication, the bounding paleosols also 364
correspond to the distal expression of other channel belts (365
366
Fig. 13). 367
6.2 How do crevasse splay sets form? 368
38
Figure 14 presents a model for splay evolution from Beaufort Group datasets that may be 369
generally applicable to flashy discharge fluvial systems in non-coal bearing strata. The 370
interpreted ephemeral and flashy nature of the fluvial system with rare plant material, 371
suggests that the banks of the channels were less stable and cohesive than channel belts 372
from organic-rich temperate settings (Smith, 1976) and therefore more prone to crevasse 373
and avulsion processes. Overbank fines typically settle out of suspension following flood 374
events, and siltstone deposits fine down paleoflow (Bridge, 2006), which represents a 375
gradual reduction in flow velocity with increasing distance from the main channel (e.g., 376
Burns et al., 2017). Unconfined crevasse splay elements deposited in distal floodplain 377
settings typically comprise siltstone to very fine-grained sandstone (378
379 Fig. 14). Each thin splay element appears to thicken slightly at different positions along 380
section, which supports an interpretation of compensational stacking of individual splays in 381
a crevasse splay set (382
39
383 Fig. 14A). Deposits from crevasse splays differ from frontal splays in adjacent strata 384
described by Gulliford et al. (2014) as their proximal deposits are much thinner (< 2 m thick), 385
they are less laterally extensive, and have paleocurrent indicators oblique to the main 386
paleochannel (387
388 Fig. 14B). Moderate maturity paleosols developed during periods or in areas of reduced 389
40
overbank flooding (390
391
Fig. 13A) as the low sedimentation rate prevented introduction of clastic sediment. 392
In order for crevasse splay sets to develop, the parent channel belt and the crevasse 393
location must stay relatively fixed during a period of multiple flood events. In crevasse- and 394
avulsion-dominated systems, fluvial channels are highly mobile, meaning that the 395
probability of future re-avulsion into (and out of) the area is high. Sedimentation rate affects 396
41
the frequency of avulsion, which takes place when the channel belt aggrades to a critical 397
height above the floodplain (Heller and Paola, 1996; Peakall, 1998). Once filled by splay 398
deposits, there is no accommodation for additional material between the floodplain surface 399
and the location of crevasse splay formation. Therefore, we suggest that a splay complex 400
can only form with the combination of low avulsion frequency and high accommodation. In 401
meandering systems, the area of the floodplain most prone to crevasse splay deposits, the 402
outer bank, is also most susceptible to erosion during bend migration. 403
6.3 Comparison with other splay datasets 404
The well exposed sections from the Great Escarpment (Supplementary Figs. 1, 2) provide a 405
database of 154 proximal and distal crevasse splay elements (for definition of proximal and 406
distal crevasse splay elements, q.v. Burns et al. [2017]). Individual proximal crevasse splay 407
elements range in thickness between 0.3 and 1.9 m and individual distal crevasse splay 408
elements range between 0.05 and 1.05 m (Supplementary Table 2). Crevasse splay element 409
geometric data from the published literature indicate that most splay elements have an 410
aspect ratio ranging between 20:1 and 4260:1 (Coleman, 1969; Stear, 1983; O'Brien and 411
Wells, 1986; Smith, 1987; Smith et al., 1989; Jordaan, 1990; Bristow et al., 1999; Morozova 412
and Smith, 2000; Farrell, 2001; Stouthamer, 2001; Gulliford et al., 2014; Burns et al., 2017) 413
(Table 3). Therefore, given that preserved mean splay thickness from the Abrahamskraal 414
Formation is 1.1 m (proximal), their lateral extent could range from 22-4686 m. However, 415
previous authors have not taken a hierarchical approach, and some examples are likely 416
crevasse splay sets. 417
42
A comparison between published crevasse splay geometry and architecture from 418
modern and ancient rivers is shown in Table 3. Some caution should be exercised when 419
comparing modern (uncompacted) and ancient (lithified) splay thickness, although the silt- 420
and sand-rich nature of these deposits mean that there will not be substantive compaction. 421
Most of these studies relate to splays from temperate climates (e.g., Coleman, 1969; Smith 422
et al., 1989; Bristow et al., 1999), preserving elongate, lobate and progradational profiles 423
that thin and fine distally. Similar geometries have also been observed in splays deposited 424
under ephemeral and flashy discharge conditions, from Clarence River (O'Brien and Wells, 425
1986) and from this study. Stage I perennial splays defined by Smith et al. (1989) from the 426
Saskatchewan River are described as small (< 1 km2) and lobate following initial overbank 427
flooding, and comparable in scale and geometry to the ephemeral crevasse splay elements 428
aヴﾗﾏ Kﾉｷヮﾆヴ;;ﾉ ;ﾐS Bﾉﾗﾏげゲ F;ヴﾏゲく There is no significant change between thicknesses of 429
these perennial and ephemeral-flashy deposits, and the formation of Stage I, II and III splays 430
defined by Smith et al. (1989) do not appear to be constrained to specific climatic conditions 431
(Table 3). However, there appears to be greater lateral variability in crevasse splay deposits 432
from ephemeral systems (Figs. 6-11), interpreted to be the result of more extreme 433
fluctuations in discharge regime, which may intensify channel avulsion behavior (Peakall, 434
1998). The process of crevasse splay formation may differ slightly between perennial and 435
ephemeral environments, with more common overtopping of channels and increased 436
overland flow in ephemeral systems. The density and type of floodplain vegetation may also 437
be an important variable, particularly in the development of crevasse channels, and the 438
amount of floodplain material entrainment during overbanking. 439
6.4 Splay elements and splay sets as part of the distributive fluvial system model 440
43
Almost one third (31%) of channel-belts analyzed in this study (n = 98) are incised directly 441
into floodplain mudstone (442
443
44
Fig. 13B) rather than into proximal overbank deposits (444
445 Fig. 13C), consistent with incisional avulsion (Hajek and Edmonds, 2014). However, the 446
majority of Beaufort channel-belts (69%) are underlain by crevasse splay elements and 447
crevasse splay sets. This style of organization has been interpreted to indicate 448
progradational avulsion (Slingerland and Smith, 2004; Jones and Hajek, 2007; Hajek and 449
Edmonds, 2014), in which repeated splay events eventually lead to a full channel avulsion. 450
45
The distributive fluvial system (DFS) paradigm refers to deposits from channels and 451
floodplains that radiate from an apex, forming a fan-shaped system (Geddes, 1960; 452
Stanistreet and McCarthy, 1993; Nichols and Fisher, 2007; Weissmann et al., 2010; 453
Weissmann et al., 2011). Not all DFSs flow to the ocean; some terminate on alluvial plains, 454
feed into axial rivers, or drain into lakes (Weissmann et al., 2010). Ancient successions 455
interpreted as DFSs include the Morrison Formation of the Colorado Plateau area (Owen et 456
al., 2017) and the lower Beaufort Group (Wilson et al., 2014; Gulliford et al., 2014). 457
In proximal DFS settings, Nichols and Fisher (2007) described the main floodplain 458
components as consisting of abandoned channel deposits. In this model, overbank deposits 459
comprising extensive floodplain mudstones and sheet sands (i.e., splays) are only preserved 460
in the medial zones. Distal to terminal settings are also typically delineated by an abundance 461
of thin, isolated sheet sandstones with small, poorly channelized deposits and an increased 462
proportion of floodplain to channel belt facies, associated with a downstream decrease in 463
the size of the channel system as sedimentation rates decrease (Nichols and Fisher 2007). 464
This would imply that overbank deposits will continue to fine away from the active channel 465
belt to a point where they are not inundated by sediment-charged floodwaters, enabling 466
paleosols to develop. This is consistent with findings from the lower Beaufort Group where 467
deposition of splays increases in the medial to distal reaches of the DFS (Gulliford et al., 468
2014), and crevasse splay elements are more likely to stack into crevasse splay sets (i.e., at 469
Sutherland East). The progradational trend of crevasse splays and splay sets beneath 470
channel belts) may correspond to the progradation of the DFS (Gulliford et al., 2014), which 471
would then predict a greater abundance of crevasse splay elements and crevasse splay sets 472
to be deposited within the distal to medial reaches of the advancing DFS, and a stratigraphic 473
change from predominantly frontal to crevasse splays (Gulliford et al. 2014). This 474
46
stratigraphic motif may therefore be an additional criterion for recognition of DFSs in the 475
ancient record. 476
7. Conclusions 477
Crevasse splay deposits are important components of fluvial-overbank successions, 478
particularly in aggradational mud-rich systems such as the lower Beaufort Group where they 479
form up to 12% of the total 810 m studied section. Locally, individual crevasse splay 480
elements stack to form crevasse splay sets up to 4 m thick. There is evidence for both 481
incisional and progradational styles of avulsion. Incisional avulsion is recognized by channel 482
belt deposits directly overlying floodplain fines. However, the great majority (69%) of 483
crevasse deposits are characterized by a gradual upwards increase in splay deposition below 484
47
a channel belt sandstone and interpreted to represent progradation(485
486
Fig. 13C). The dominance of progradational styles may be characteristic of a distributive 487
fluvial system, rather than a trunk river system. 488
Overall, the thicknesses of splay deposits are comparable in different climatic 489
settings; however, the major fluctuations in flood hydrograph and increased frequency of 490
splay deposits associated with ephemeral systems such as the Beaufort Group is reflected in 491
48
the wider range of crevasse splay widths (and aspect ratios) in deposits from small 492
ephemeral streams, compared to those from larger perennial rivers. Crevasse splay deposits 493
have previously been understudied in distributive fluvial systems. Splay abundance, 494
sedimentological characteristics and stacking patterns all play an important role in 495
identifying DFS zones (i.e., proximal, medial, distal, terminal), as well as determining 496
avulsion frequency and floodbasin accommodation. 497
Acknowledgements 498
We acknowledge Chevron Australia Pty Ltd for funding the Beaufort Project. Special thanks 499
to Tobias Payenberg, Anne Powell and Brian Willis. John Kavanagh, Laura Fielding, Ashley 500
Clarke, Janet Richardson and Andrew Wilson are thanked for their support in the field. We 501
also acknowledge the many landowners in the Sutherland area for kindly granting land 502
access. In addition, this manuscript has significantly benefited from detailed and 503
constructive feedback from Frank Ethridge, an anonymous reviewer and the editor, Jasper 504
Knight. 505
49
References 506
Abels, H.A., Kraus, M.J., Gingerich, P.D., 2013. Precession-scale cyclicity in the fluvial lower 507
Eocene Willwood Formation of the Bighorn Basin, Wyoming (USA). Sedimentology 508
60, 1467Ъ1483. 509
Allen, J.R.L., 1973. A classification of climbing-ripple cross-lamination. Journal of the 510
Geological Society 129, 537Ъ541. 511
Allen, J.R.L., 1983. Studies in fluviatile sedimentation: Bars, bar-complexes and sandstone 512
sheets (low-sinuosity braided streams) in the Brownstones (L. Devonian), Welsh 513
Borders. Sedimentary Geology 33, 237Ъ293. 514
Ambrose, W.A., Lakshminarasimhan, S., Holtz, M.H., Nunez-Lopez, V., Hovorka, S.D., 515
Duncan, I., 2008. Geologic factors controlling CO2 storage capacity and permanence: 516
case studies based on experience with heterogeneity in oil and gas reservoirs applied 517
to CO2 storage. Environmental Geology 54, 1619Ъ1633. 518
Ayers Jr, W.B., 2002. Coalbed gas systems, resources, and production and a review of 519
contrasting cases from the San Juan and Powder River basins. AAPG Bulletin 86, 520
1853Ъ1890. 521
Blakey, R.C., Gubitosa, R., 1984. Controls of sandstone body geometry and architecture in 522
the Chinle Formation (Upper Triassic), Colorado Plateau. Sedimentary Geology 38, 523
51Ъ86. 524
Bridge, J.S., 1984. Large-scale facies sequences in alluvial overbank environments. Journal of 525
Sedimentary Petrology 54, 583Ъ588. 526
Bridge, J.S., 2003. Rivers and floodplains: Forms, processes, and sedimentary record. 527
Blackwell Science, Oxford 491 pp. 528
50
Bridge, J.S., 2006. Fluvial facies models: Recent developments. In: Posamentier, H.W., 529
Walker, R.G. (Eds.), Facies Models Revisited. SEPM Special Publication 84, Tulsa, OK., 530
pp. 85Ъ170. 531
Bridge, J.S., Tye, R.S., 2000. Interpreting the dimensions of ancient fluvial channel bars, 532
channels, and channel belts from wireline-logs and cores. AAPG Bulletin 84, 533
1205Ъ1228. 534
Bristow, C.S., Skelly, R.L., Ethridge, F.G., 1999. Crevasse splays from the rapidly aggrading, 535
sand-bed, braided Niobrara River, Nebraska: effect of base-level rise. Sedimentology 536
46, 1029Ъ1048. 537
Burns, C.E., Mountney, N.P., Hodgson, D.M., Colombera, L., 2017. Anatomy and dimensions 538
of fluvial crevasse-splay deposits: Examples from the Cretaceous Castlegate Sandstone and 539
Neslen Formation, Utah, USA. Sedimentary Geology, 351, 21-35. 540
Coleman, J.M., 1969. Brahmaputra River: channel processes and sedimentation. 541
Sedimentary Geology 3, 129Ъ239. 542
Colombera, L., Mountney, N.P., McCaffrey, W.D., 2012. A relational database for the 543
digitization of fluvial architecture concepts and example applications. Petroleum 544
Geoscience 18, 129Ъ140. 545
Colombera, L., Mountney, N.P., McCaffrey, W.D., 2013. A quantitative approach to fluvial 546
facies models: Methods and example results. Sedimentology 60, 1526Ъ1558. 547
Council for Geoscience, 1983. Geological map 3220 Sutherland Council for Geoscience, 548
Pretoria. 549
Dubiel, R.F., 1987. Sedimentology of the Upper Triassic Chinle Formation southeastern Utah: 550
Paleoclimatic implications. Journal of the Arizona-Nevada Academy of Science 22, 551
35Ъ45. 552
51
Ethridge, F.G., Jackson, T.J., Youngberg, A.D., 1981. Floodbasin sequence of a fine-grained 553
meander belt subsystem: the coal-bearing Lower Wasatch and Upper Fort Union 554
Formations Southern Powder River Basin, Wyoming. In: Ethridge, F.G., Flores, R.M. 555
(Eds.). Recent and ancient nonmarine depositional environments に Models for 556
exploration: SEPM Special Publication 31, 191Ъ209. 557
Farrell, K.M., 2001. Geomorphology, facies architecture, and high-resolution, non-marine 558
sequence stratigraphy in avulsion deposits, Cumberland Marshes, Saskatchewan. 559
Sedimentary Geology 139, 93Ъ150. 560
Fielding, C.R., 2006. Upper flow regime sheets, lenses and scour fills: Extending the range of 561
architectural elements for fluvial sediment bodies. Sedimentary Geology 190, 562
227Ъ240. 563
Fielding, C.R., Allen, J.P., Alexander, J., Gibling, M.R., 2009. Facies model for fluvial systems 564
in the seasonal tropics and subtropics. Geology, 37, 623-626. 565
Fielding, C.R., Ashworth, P.J., Best, J.L., Prokocki, E.W., Sambrook Smith., S.H., 2012. 566
Tributary, distributary and other fluvial patterns: What really represents the norm in 567
the continental rock record?. Sedimentary Geology 261, 15-32. 568
Fisher, J.A., Krapf, C.B.E., Lang, S.C., Nichols, G.J., Payenberg, T.H.D., 2008. Sedimentology 569
and architecture of the Douglas Creek terminal splay, Lake Eyre, central Australia. 570
Sedimentology 55, 1915Ъ1930. 571
Flint, S.S., Hodgson, D.M., Sprague, A.R., Brunt, R.L., Van der Merwe, W.C., Figueiredo, J., 572
Prélat, A., Box, D., Di Celma, C., Kavanagh, J.P., 2011. Depositional architecture and 573
sequence stratigraphy of the Karoo basin floor to shelf edge succession, Laingsburg 574
depocentre, South Africa. Marine and Petroleum Geology 28, 658Ъ674. 575
52
Flores, R.M., 1983. Basin facies analysis of coal-rich Tertiary fluvial deposits, northern 576
Powder River Basin, Montana and Wyoming. In: Collinson, J.D., Lewin, J. (Eds.), 577
Modern and ancient fluvial systems: International Association of Sedimentologists 578
Special Publication 6, Oxford, pp. 499Ъ515. 579
Ford, G.L., Pyles, D.R., 2014. A hierarchical approach for evaluating fluvial systems: 580
Architectural analysis and sequential evolution of the high net-sand content middle 581
Wasatch Formation, Uinta Basin, Utah. AAPG Bulletin 98, 1273Ъ1304. 582
Friend, P.F., 1983. Towards the field classification of alluvial architecture or sequence. 583
Modern and ancient fluvial systems. International Association of Sedimentologists 584
Special Publication Oxford 345 pp. 585
Geddes, A., 1960. The alluvial morphology of the Indo-Gangetic Plain: Its mapping and 586
geographical significance. Transactions and Papers (Institute of British Geographers) 587
28, 253-276. 588
Gouw, M.J.P., Berendsen, H.J.A., 2007. Variability of channel belt dimensions and the 589
consequences for alluvial architecture: observations from the Holocene Rhine-590
Meuse delta (The Netherlands) and Lower Mississippi Valley (USA). Journal of 591
Sedimentary Research 77, 124Ъ138. 592
Gulliford, A.R., 2014. Controls on river and overbank processes in an aggradation-dominated 593
system: Permo-Triassic Beaufort Group, South Africa. PhD Thesis, University of 594
Manchester, UK, 361 pp. 595
Gulliford, A.R., Flint, S.S., Hodgson, D.M., 2014. Testing applicability of models of distributive 596
fluvial systems or trunk rivers in ephemeral systems: Reconstructing 3-D fluvial 597
architecture in the Beaufort Group, South Africa. Journal of Sedimentary Research 598
84, 1147Ъ1169. 599
53
Hajek, E.A., Edmonds, D.A., 2014. Is river avulsion style controlled by floodplain 600
morphodynamics? Geology 42, 199Ъ202. 601
Hartley, A.J., Weissmann, G.S., Nichols, G.J., Scuderi, L.A., 2010a. Fluvial form in modern 602
continental sedimentary basins: Distributive fluvial systems: REPLY. Geology 38, 231-603
231. 604
Hartley, A.J., Weissmann, G.S., Nichols, G.J., Warwick, G.L., 2010b. Large distributive fluvial 605
systems: characteristics, distribution, and controls on development. Journal of 606
Sedimentary Research 80, 167-183. 607
Hasiotis, S.T., Van Wagoner, J.C., Demko, T.M., Wellner, R.W., Jones, C.R., Hill, R.E., 608
McCrimmon, G.G., Feldman, H.R., Drzewiecki, P.A., Patterson, P. and Donovan, A.D. 609
(2002). Continental ichnology: using terrestrial and freshwater trace fossils for 610
environmental and climatic interpretations. SEPM Short Course, 51. 611
Heller, P.L., Paola, C., 1996. Downstream changes in alluvial architecture: an exploration of 612
controls on channel-stacking patterns. Journal of Sedimentary Research 66, 297-306. 613
Horne, J.C., Ferm, J.C., Caruccio, F.T., Baganz, B.P., 1978. Depositional models in coal 614
exploration and mine planning in Appalachian region. AAPG Bulletin 62, 2379Ъ2411. 615
Iwere, F.O., Moreno, J.E., Apaydin, O.G., 2006. Numerical Simulation of Thick Tight Fluvial 616
Sands. SPE Reservoir Evaluation and Engineering 9, 374Ъ381. 617
Jenson, M.A., Pedersen, G.K., 2010. Architecture of vertically stacked fluvial deposits, Atane 618
Formation, Cretaceous, Nuussuaq, central West Greenland. Sedimentology 57, 619
1280Ъ1314. 620
Jerrett, R.M., Hodgson, D.M., Flint, S.S., Davies, R.C., 2011. Control of Relative Sea Level and 621
Climate on Coal Character in the Westphalian C (Atokan) Four Corners Formation, 622
Central Appalachian Basin, USA. Journal of Sedimentary Research 81, 420Ъ445. 623
54
Johnson, M.R., Van Vuuren, C.J., Visser, J.N.J., Cole, D.I., Wickens, H.D.V., Christie, A.D.M., 624
Roberts, D.L., 1997. The foreland Karoo Basin, South Africa. In: Selley, R.C. (Ed.), 625
Sedimentary Basins of the World. African Basins. Elsevier Science, Amsterdam, pp. 626
269Ъ317. 627
Jones, H.L., Hajek, E.A., 2007. Characterizing avulsion stratigraphy in ancient alluvial 628
deposits. Sedimentary Geology 202, 124Ъ137. 629
Jopling, A.V., Walker, R.G., 1968. Morphology and origin of ripple-drift cross-lamination, 630
with examples from the Pleistocene of Massachusetts. Journal of Sedimentary 631
Research 38, 971Ъ984. 632
Jordaan, M.J., 1990. Basin analysis of the Beaufort Group in the western part of the Karoo 633
Basin. PhD Thesis, University of the Orange Free State, South Africa, 283 pp. 634
Jørgensen, P.J., Fielding, C.R., 1996. Facies architecture of alluvial floodbasin deposits: three-635
dimensional data from the Upper Triassic Callide Coal Measures of east-central 636
Queensland, Australia. Sedimentology 43, 479Ъ495. 637
Kraus, M.J., 1987. Integration of channel and floodplain suites, II. Vertical relations of alluvial 638
paleosols. Journal of Sedimentary Research 57, 602Ъ612. 639
Kraus, M.J., Aslan, A., 1993. Eocene hydromorphic paleosols: significance for interpreting 640
ancient floodplain processes. Journal of Sedimentary Research 63, 453-463. 641
Kraus, M.J., Wells, T.M., 1999. Recognizing avulsion deposits in the ancient stratigraphical 642
record. In: Smith, N.D., Rogers, J. (Eds.), Fluvial Sedimentology VI, Special Publication 643
International Association of Sedimentologists 28, 251Ъ268. 644
Kraus, M.J., Hasiotis, S.T., 2006. Significance of different modes of rhizolith preservation to 645
interpreting paleoenvironmental and paleohydrologic settings: examples from 646
55
Paleogene paleosols, Bighorn Basin, Wyoming, USA. Journal of Sedimentary 647
Research 76, 633Ъ646. 648
Leeder, M.R., 1975. Pedogenic carbonates and flood sediment accretion rates: a 649
quantitative model for alluvial arid-zone lithofacies. Geological Magazine 112, 650
257Ъ270. 651
Miall, A.D., 1985. Architectural-element analysis Ъ a new method of facies analysis applied 652
to fluvial deposits. Earth-Science Reviews 22, 261Ъ308. 653
Miall, A.D., 1988. Architectural elements and bounding surfaces in fluvial deposits: anatomy 654
of the Kayenta Formation (Lower Jurassic), south-west Colorado. Sedimentary 655
Geology 55, 233Ъ262. 656
Miall, A.D., 1996. The geology of fluvial deposits. Springer-Verlag, New York, 582 pp. 657
Mjøs, R., Walderhaug, O., Prestholm, E., 1993. Crevasse splay sandstone geometries in the 658
Middle Jurassic Ravenscar Group of Yorkshire, UK. In: Marzo, M., Puigdefábregas, C. 659
(Eds.), Alluvial Sedimentation, International Association of Sedimentologists, Special 660
Publication 17. John Wiley and Sons, Oxford, 184 pp. 661
Mohrig, D., Heller, P.L., Paola, C., Lyons, W.J., 2000. Interpreting avulsion process from 662
ancient alluvial sequences: Guadalope-Matarranya system (northern Spain) and 663
Wasatch Formation (western Colorado). Geological Society of America Bulletin 112, 664
1787-1803. 665
Morozova, G.S., Smith, N.D., 2000. Holocene avulsion styles and sedimentation patterns of 666
the Saskatchewan River, Cumberland Marshes, Canada. Sedimentary Geology 130, 667
81Ъ105. 668
Nanson, G.C., Croke, J.C., 1992. A genetic classification of floodplains. Geomorphology 4, 669
459Ъ486. 670
56
Nichols, G.J., 2009. Sedimentology and stratigraphy. John Wiley and Sons, Oxford, 432 pp. 671
Nichols, G.J., Fisher, J.A., 2007. Processes, facies and architecture of fluvial distributary 672
system deposits. Sedimentary Geology 195, 75Ъ90. 673
Owen, A., Nichols, G.J., Hartley, A.J., Weissmann, G.S., 2017. Vertical trends within the 674
prograding Salt Wash distributive fluvial system, SW United States. Basin Research 675
29, 64-80. 676
O'Brien, P.E., Wells, A.T., 1986. A small, alluvial crevasse splay. Journal of Sedimentary 677
Research 56, 876Ъ879. 678
Pashin, J.C., 1998. Stratigraphy and structure of coalbed methane reservoirs in the United 679
States: an overview. International Journal of Coal Geology 35, 209Ъ240. 680
Payenberg, T.H.D., Reilly, M.R.W., Lang, S.C., Kassan, J., 2004. Sedimentary processes in an 681
ephemeral river and terminal splay complex, Lake Eyre, Central Australia - 682
observation from the February 2003 flooding event. AAPG Search and Discovery 683
Article #90034. 684
Payenberg, T.H.D., Willis, B.J., Bracken, B., Posamentier, H.W., Pyrcz, M.J., Pusca, V., 685
Sullivan, M.D., 2011. Revisiting the subsurface classification of fluvial sandbodies. 686
AAPG Search and Discovery Article #90124. 687
Peakall, J., 1998. Axial river evolution in response to half-graben faulting: Carson River, 688
Nevada, USA. Journal of Sedimentary Research 68, 788-799. 689
Pranter, M.J., Cole, R.D., Panjaitan, H., Sommer, N.K., 2009. Sandstone-body dimensions in a 690
lower coastal-plain depositional setting: Lower Williams Fork Formation, Coal 691
Canyon, Piceance Basin, Colorado. AAPG Bulletin 93, 1379Ъ1401. 692
Pysklywec, R.N., Mitrovica, J.X., 1999. The role of subduction-induced subsidence in the 693
evolution of the Karoo Basin. The Journal of Geology 107, ヱヵヵЪ164. 694
57
Sambrook Smith, G.H., Best, J.L., Ashworth, P.J., Fielding, C.R., Goodbred, S.L., Prokocki, 695
E.W., 2010. Fluvial form in modern continental sedimentary basins: Distributive 696
fluvial systems: Comment. Geology 38, 230-230. 697
Shanley, K.W., Cluff, R.M., Robinson, J.W., 2004. Factors controlling prolific gas production 698
from low-permeability sandstone reservoirs: Implications for resource assessment, 699
prospect development, and risk analysis. AAPG Bulletin 88, 1083Ъ1121. 700
Singh, H., Parkash, B., Gohain, K., 1993. Facies analysis of the Kosi megafan deposits. 701
Sedimentary Geology 85, 87Ъ113. 702
Slingerland, R., Smith, N.D., 2004. River avulsions and their deposits. Annual Review of Earth 703
and Planetary Sciences 32, 257Ъ285. 704
Smith, D.G., 1976. Effect of vegetation on lateral migration of anastomosed channels of a 705
glacier meltwater river. Geological Society of America Bulletin 87, 857Ъ860. 706
Smith, N.D., Cross, T.A., Dufficy, J.P., Clough, S.R., 1989. Anatomy of an avulsion. 707
Sedimentology 36, 1Ъ23. 708
Smith, R.M.H., 1987. Morphology and depositional history of exhumed Permian point bars 709
in the southwestern Karoo, South Africa. Journal of Sedimentary Petrology 57, 710
19Ъ29. 711
Smith, R.M.H., 1990a. Alluvial Paleosols and Pedofacies Sequences in the Permian Lower 712
Beaufort of the Southwestern Karoo Basin, South-Africa. Journal of Sedimentary 713
Petrology 60, 258Ъ276. 714
Smith, R.M.H., 1990b. A review of stratigraphy and sedimentary environments of the Karoo 715
Basin of South Africa. Journal of African Earth Sciences 10, 117Ъ137. 716
Smith, R.M.H., 1993a. Sedimentology and Ichnology of Floodplain Paleosurfaces in the 717
Beaufort Group (Late Permian), Karoo Sequence, South Africa. Palaios 8, 339Ъ357. 718
58
Smith, R.M.H., 1993b. Vertebrate taphonomy of Late Permian floodplain deposits in the 719
southwestern Karoo Basin of South Africa. Palaios 8, 45Ъ67. 720
Stear, W.M., 1978. Sedimentary structures related to fluctuating hydrodynamic conditions 721
in flood plain deposits of the Beaufort Group near Beaufort West, Cape. Transactions 722
of the Geological Society of South Africa 81, 393Ъ399. 723
Stear, W.M., 1980. The sedimentary environment of the Beaufort Group Uranium Province 724
in the vacinity of Beaufort West, South Africa. PhD Thesis, University of Port 725
Elizabeth, South Africa, 257 pp. 726
Stear, W.M., 1983. Morphological characteristics of ephemeral stream channel and 727
overbank splay sandstone bodies in the Permian lower Beaufort Group, Karoo Basin, 728
South Africa. In: Collinson, J.D., Lewin, J. (Eds.), Modern and ancient fluvial systems: 729
International Association of Sedimentologists Special Publication 6, Oxford, pp. 730
405Ъ420. 731
Stouthamer, E., 2001. Sedimentary products of avulsions in the Holocene RhineにMeuse 732
delta, The Netherlands. Sedimentary Geology 145, 73Ъ92. 733
Stuart, J.Y., Mountney, N.P., McCaffrey, W.D., Lang, S., Collinson, J.D., 2014. Prediction of 734
channel connectivity and fluvial style in the flood basin successions of the Upper 735
Permian Rangal Coal Measures (Queensland). AAPG Bulletin 98, 191Ъ212. 736
Tankard, A., Welsink, H., Aukes, P., Newton, R., Stettler, E., 2009. Tectonic evolution of the 737
Cape and Karoo basins of South Africa. Marine and Petroleum Geology 26, 738
1379Ъ1412. 739
Taylor, A.M., Goldring, R., 1993. Description and analysis of bioturbation and ichnofabric. 740
Journal of the Geological Society 150, 141Ъ148. 741
59
Turner, B.R., 1981. Possible origin of low angle cross-strata and horizontal lamination in 742
Beaufort Group sandstones of the southern Karoo Basin. Transactions of the 743
Geological Society of South Africa 84, 193Ъ197. 744
Tyler, N., Ethridge, F.G. (Eds.), 1983. Fluvial architecture of Jurassic uranium-bearing 745
sandstones, Colorado Plateau, western United States. Modern and ancient fluvial 746
systems. International Association of Sedimentologists Special Publication Oxford, 747
575 pp. 748
van Toorenenburg, K.A., Donselaar, M.E., Noordijk, N.A., Weltje, G.J., 2016. On the origin of 749
crevasse-splay amalgamation in the Huesca fluvial fan (Ebro Basin, Spain): 750
Implications for connectivity in low net-to-gross fluvial deposits. Sedimentary 751
Geology 343, 156-164. 752
Walling, D.E., He, Q., 1998. The spatial variability of overbank sedimentation on river 753
floodplains. Geomorphology 24, 209Ъ223. 754
Weissmann, G.S., Hartley, A.J., Nichols, G.J., Scuderi, L.A., Olson, M., Buehler, H., Banteah, 755
R., 2010. Fluvial form in modern continental sedimentary basins: Distributive fluvial 756
systems. Geology 38, 39-42. 757
Weissmann, G.S., Hartley, A.J., Nichols, G.J., Scuderi, L.A., Olson, M.E., Buehler, H.A., 758
Massengill, L.C., 2011. Alluvial facies distributions in continental sedimentary 759
basinsねdistributive fluvial systems. In: Davidson, S.K., Leleu, S., North, C.P. (Eds.), 760
From River to Rock Record: The Preservation of Fluvial Sediments and their 761
Subsequent Interpretation. SEPM Special Publication 97, Tulsa, OK., pp. 327-355. 762
Wickens, H., 1994. Basin floor fan building turbidites of the Southwestern Karoo Basin, 763
Permian Ecca Group, South Africa. PhD Thesis, University of Port Elizabeth, South 764
Africa, 233 pp. 765
60
Willis, B.J., Behrensmeyer, A.K., 1994. Architecture of Miocene overbank deposits in 766
northern Pakistan. Journal of Sedimentary Research 64, 60Ъ67. 767
Wilson, A., Flint, S.S., Payenberg, T.H.D., Tohver, E., Lanci, L., 2014. Architectural styles and 768
sedimentology of the fluvial lower Beaufort Group, Karoo Basin, South Africa. Journal 769
of Sedimentary Research 84, 326Ъ348. 770
Wright, V.P., Marriott, S.B., 1993. The sequence stratigraphy of fluvial depositional systems: 771
the role of floodplain sediment storage. Sedimentary Geology 86, 203Ъ210. 772
Wright, V.P., Taylor, K.G., Beck, V.H., 2000. The paleohydrology of Lower Cretaceous 773
seasonal wetlands, Isle of Wight, Southern England. Journal of Sedimentary Research 774
70, 619Ъ632. 775
61
Figure Captions 776
62
777
Fig. 1. GoogleEarthTM Landsat image of the main localities, marked by the white dots, the 778
Great Escarpment and the town of Sutherland. The satellite image is overlain by a simplified 779
geological map of the area (Council for Geoscience, 1983). Localities (all listed using UTM 780
WGS84) include: Blom's Farm [468920.64 m E, 6422756.7 m N, 34H]; Boschmanshoek 781
[467594.89 m E, 6391162.5 m S, 34H]; Brandwacht [470425.59 m E, 6439131.2 m N, 34H]; 782
Klipkraal Farm [471280.65 m E, 6417284.1 m N, 34H]; Klipkraal Lateral [474445.12 m E, 783
6414923.2 m N, 34H]; Müller Canyon [465094.72 m E, 6396063.1 m N, 34H]; Oliviersburg 784
West [473316.4 m E, 6389355.8 m N, 34H]; Perdekop [464577.65 m E, 6390132.1 m N, 785
34H]; Sutherland East [468682.08 m E, 6416176 m N, 34H]; Verlatenkloof [465716.15 m E, 786
6401131 m N, 34H]. Inset: map of southern Africa, highlighting the study area, within the 787
SW Karoo Basin, Northern Cape Province. 788
63
789
Fig. 2. Lithostratigraphy of the SW Karoo Basin. (A) Cape and Karoo Supergroups highlighting 790
the Abrahamskraal Formation, lower Beaufort Group, redrawn after Wickens (1994) and 791
Flint et al. (2011). (B) Schematic illustration of the Abrahamskraal Formation exposed on the 792
Great Escarpment (Verlatenkloof and Müller Canyon に refer to Figure 1 for GPS coordinates) 793
and in the Sutherland area (Klipkraal, S┌デｴWヴﾉ;ﾐS E;ゲデが ;ﾐS Bﾉﾗﾏげゲ F;ヴﾏぶ. 794
795
64
Fig. 3. Sandstone to mudstone proportions and architectural elements in sedimentary logs 796
from (A) Verlatenkloof, (B) Müller Canyon. See Figure 1 for locations and Supplementary 797
Spreadsheet 1 for supporting dataset. 798
799
Fig. 4. Mudstone facies from the Abrahamskraal Formation. (A) Downward-branching fossil 800
root structures (rhizoliths) preserved as alteration-reduction halos (Hasiotis et al., 2002) in a 801
paleosol (Fp), (B) Poorly sorted purple siltstone (Fpu), overlying Fbg, above Fcs. Dashed 802
white lines show a sharp lower bounding surface, (C) Bright green massive mudstone (Fbg), 803
mottled purple. Facies is flanked by sharp-topped, erosive based coarse-grained siltstone to 804
very fine-grained sandstone. Fbg is typically lenticular, (D) Poorly sorted, green-gray-blue, 805
coarse-grained siltstone (Fggb) forming laterally extensive sheets. (E) Laminated organic-rich 806
dark gray mudstone (Fgl). Fgl facies is thinly-bedded (< 5 cm thick) flanking a possible tuff 807
deposit (~1 cm thick). (F) Thinly-bedded coarse-grained siltstone to very fine-grained 808
sandstone (Fcs), with sharp upper and lower bounding surfaces from Sutherland East. Fcs 809
facies is gray colored, weathering orange. 810
65
811
Fig. 5. Idealized sedimentary log showing relative proportions of overbank deposits 812
observed throughout the study area, focusing on floodplain lake deposits (FFL), floodplain 813
fines element (FF) and splay element (SS) comprising proximal to distal splay and crevasse 814
channel deposits. 815
66
816
67
Fig. 6. Overview map and representative photographs of an interpreted crevasse splay 817
element. (A) Overview map of Klipkraal Farm, near Sutherland (818
819
Fig. 1). The location of this map is given by the blue rectangle in the inset map. Red dots 820
denote sedimentary log localities. (B) Photograph of log locality KK-30. Refer to Figure 7 for 821
each sedimentary log. (C) Photograph of log locality KK-29. (D) Photograph of log locality KK-822
28. (E) Photograph of log locality KK-27. 823
68
824
69
Fig. 7. Sedimentary logs through an interpreted crevasse splay element 1.5 m thick at 825
Klipkraal Farm, near Sutherland (826
827
Fig. 1). Paleocurrents from ripple and climbing-ripple cross-lamination are to the north. A 828
laterally extensive paleosol is present < 1 m above the top of the crevasse splay sandstone. 829
Erosional base of splay is highlighted in blue. Inset: Photograph of climbing-ripple cross-830
lamination. Refer to Figure 6 for locality map. 831
70
832
71
Fig. 8. Overview map and representative photographs of an interpreted crevasse splay 833
element. ふAぶ O┗Wヴ┗ｷW┘ ﾏ;ヮ ﾗa Bﾉﾗﾏげゲ F;ヴﾏが ﾐW;ヴ S┌デｴWヴﾉ;ﾐS (834
835
Fig. 1). The location of this map is given by the blue rectangle in the inset map. Red dots 836
denote sedimentary log localities. (B) Photograph of log locality BL-20, comprising three 837
coarsening-upward packages. Refer to Figure 9 for each sedimentary log. (C) Photograph of 838
log locality BL-19. (D) Photograph of log locality BL-18. 839
72
840
73
Fig. 9. Sedimentary logs through a a < 1.5 m thick very fine-grained sandstone crevasse 841
splay element ;デ Bﾉﾗﾏげゲ F;ヴﾏ ふ842
843
Fig. 1). Climbing ripple cross-lamination is overlain by ripple cross-lamination, with 844
paleoflow north-northeastward. Erosional base of splay is highlighted in blue. Refer to 845
Figure 8 for locality map. 846
74
847
75
Fig. 10. Overview map and representative photographs of an interpreted crevasse splay 848
set. (A) Overview map of Sutherland East (849
850
Fig. 1), near Sutherland (3 x vertical exaggeration applied with perspective view). The 851
location of this map is given by the blue rectangle in the inset map. Red dots denote 852
sedimentary log localities. (B) Crevasse splay elements adjacent to log locality SE-8. Refer to 853
Figure 11 for each sedimentary log. (C) Photograph of log locality SE-1, showing ripple to 854
climbing ripple cross-laminated very fine- to fine-grained sandstone sheets separated 855
vertically by thin siltstone deposits. (D) Photograph of log locality SE-2 displaying laterally 856
extensive crevasse splay elements. (E) Photograph of log locality SE-3. A laterally extensive 857
nodular horizon can be observed beneath a tabular very fine-grained sandstone. (F) 858
Photograph of log locality SE-4, depicting an interpreted crevasse splay set bound above by 859
a paleosol and below by a nodular horizon. (G) Photograph of log locality SE-5, showing 860
stacked crevasse splay elements above a nodular horizon. 861
76
862
77
Fig. 11. Detailed photo panel interpretation of crevasse splay elements from Sutherland 863
East (864
865
Fig. 1), with sedimentary log positions and correlated surfaces (E to Y) overlain. Paleoflow 866
from ripple cross-laminated very fine-grained sandstone ranges between 010° and 079° (n = 867
9). The sedimentary facies distributions are shown on the accompanying logs. Refer to 868
Figure 10 for locality map. 869
870
78
871
872
Fig. 12. Crevasse splay hierarchical scheme. Individual crevasse splay elements stack to form 873
crevasse splay sets. Each crevasse splay set is separated by a laterally extensive paleosol. 874
79
875
Fig. 13. Crevasse products in the Abrahamskraal Formation. (A) Schematic diagram depicting 876
the formation of paleosols through time. Paleosols may represent the distal expression of 877
channel belts, developing during the equivalent time that it takes for them to form. 878
Alternatively, the paleosol may represent a major avulsion, and may be the hierarchical 879
expression of a channel belt complex, implying that the overbank fines between each 880
paleosol are potentially the hierarchical equivalent to a channel belt complex. (B) Examples 881
of channel belts that incise into distal floodplain mudstones, indicating incisional avulsion. 882
(C) Example of channel belt deposits overlying laterally extensive splay deposits, , indicative 883
of the dominant progradational avulsion style. 884
80
885 Fig. 14. Idealized crevasse splay model. (A) Compensationally stacked splay deposits infilling 886
topographic lows. Numbers relate to order of deposition. (B) Crevasse splay 3-D model 887
based on observations made in this study, modified after Bridge (2003); (i) Klipraal Farm 888
[472104 m E, 6417930 m N, 34H]; (ii) Klipkraal Farm, log KK-30 [472263 m E, 6417545 m N, 889
34H]; (iii) Sutherland East [468634 m E, 6415960 m N, 34H]; (iv) Klipkraal Farm, log KK-27 890
[472188 m E, 6417598 m N, 34H]き ふ┗ぶ Bﾉﾗﾏげゲ F;ヴﾏ, log BL-18 [469120 m E, 6423019 m N, 891
34H]. 892
893
81
Table Captions 894
Overbank facies (%) Overbank architectural elements (%) Grain size of elements
Fp 3.66, n = 48
FF
(floodplain fines)
79.78 claystone, siltstone (very fine- to medium-grained) Fpu 31.92, n = 375
Fbg 0.17, n = 8
Fggb 44.03, n = 518
Fgl 0.21, n = 7 FFL
(floodplain lake deposits) 0.21 claystone, siltstone (very fine- to medium-grained)
Fcs 9.22, n =107
SS
(proximal to distal
splay deposits and
crevasse channel)
20.01 siltstone (medium to coarse-grained), sandstone
(very fine- to fine-grained)
Sr 4.20, n = 45
Sm 5.68, n = 55
Sh 0.32, n = 3
Sl 0.43, n = 4
St 0.12, n = 1
Fp, fissile purple mudstone; Fpu, poorly-sorted purple siltstone; Fbg, bright green massive mudstone; Fggb, poorly-sorted
green-gray-blue siltstone; Fgl, laminated organic-rich dark gray mudstone; Fcs, sharp-based thinly bedded coarse siltstone or
very fine sandstone; Sr, ripple cross-laminated sandstone; Sm, massive / structureless sandstone; Sh, planar-laminated
sandstone; Sl, low angle cross-stratified sandstone; St, trough cross-stratified sandstone.
895
Table 1. Approximate overbank proportions of facies and architectural elements within the 896
Abrahamskraal Formation, averaged from sedimentary logs measured at Verlatenkloof and 897
Müller Canyon. Facies and elements are adapted from schemes by Allen (1983), Friend 898
(1983), Miall (1985, 1988, 1996) and Colombera et al. (2012, 2013). Data from channelized 899
deposits are excluded, as are ambiguous data where the exposure is too poor to accurately 900
quantify facies. Refer to Figure 4 for mudstone facies photographs. 901
Locality Architecture Sedimentary facies
Apparent
W
(m)
W*
(m)
T
(m)
Aspect
ratio*
Paleocurrent
direction
L
(m)
Klipkraal
Farm
Crevasse
splay
Very fine-grained sandstone.
Structureless to normally
graded or ripple and
climbing ripple cross-
lamination.
100 78 < 1.5 52
Observed range
347° to 042°
(n = 11)
-
Blom's
Farm
Crevasse
splay
Very fine-grained sandstone.
Abundant climbing ripple
cross-lamination overlain by
ripple cross-lamination.
190 62 < 2 31
Observed range
356° to 066°
(n = 14)
190
Sutherland
East Splay set
Ripple and climbing-ripple
cross-laminated very fine-
grained sandstone.
> 700m - 4 -
Observed range
010° to 079°
(n = 12)
-
W, width; T, maximum thickness recorded from splay (or splay set) axis; L, minimum length of crevasse splay as recorded from total
outcrop exposure. *Width corrected for paleocurrent relative to the strike of the outcrop using trigonometry (i.e., true width).
82
Table 2. Geometric data for crevasse splay deposits from the Abrahamskraal Formation. 902
Stratigraphic
section Age Climate Characteristics Width Thickness
Aspect
ratio Reference
Brahmaputra
River,
Bangladesh
Moder
n
Humid
tropic
al
mons
oon
Single and multichannel
crevasse splays
encompassing silt and fine-
grained sand. Distal margins
are sharp.
- 1 - 3 m - Coleman (1969)
Sandy Creek,
Clarence
River, New
South
Wales,
Australia
Moder
n
Humid
subtro
pical
to
tempe
rate
Ephemeral stream deposits.
Lobe-shaped in plan view.
Progradational in cross-
section, thinning marginally.
Sand-dominated, comprising
cross-lamination, ripple-cross
lamination and parallel
lamination. Interpreted by
Smith et al. (1989) to
represent Stage I splays.
< 20 m
(by ~40
m)
0.4 m 50 O'Brien and Wells
(1986)
Niobrara
River, U.S.A
- east splay Moder
n
Tempera
te
(with
swam
ps)
Coarsening-up sequences,
individual beds fine up.
Single channel, sand-
dominated crevasse splays.
200 m
(by 1000
m)
2.5 m 80
Bristow et al.
(1999) Niobrara
River, U.S.A
- west splay
150 m
(by 250
m)
1.2 m 125
Cumberland
Marshes,
Saskatchew
an, Canada
Holoce
ne to
Mod
ern
Subhumi
d,
boreal
(wetla
nds)
Smaller stage I crevasse splays
are lobate in plan view and
wedge-shaped in the
direction of progradation.
Stage II and III splays form
elliptical to elongate sets,
comprising silt to very fine-
grained sand.
200 に 500
m (area <
1 km2
to < 20
km2)
< 2 m ~100 に
250
Smith et al. (1989),
Morozova and
Smith (2000),
Farrell (2001)
Rhine-Meuse
Delta, The
Netherlands
Holoce
ne
Tempera
te
Sandstone-siltstone-claystone
sheets with a sharp to
gradational basal contact.
Stage II and III splays of
Smith et al. (1989).
100s of m 0.5 - 2 m ~50 に
600 Stouthamer (2001)
Castlegate
Sandstone
& Neslen
Formation,
Utah, U.S.A.
に proximal
splay Cretac
eous
Humid
subtropi
cal,
green
house
Fine-grained siltstone to upper
fine-grained sandstone.
Structureless sandstone and
ripple cross-laminated
sandstones dominate in
proximal settings, with
structureless and soft
sediment deformed chaotic
sandstone and siltstone in
distal settings.
75 に 676
m; mean
278 m
(by 55 に
189 m;
mean 129
m)
1.0 に 3.7
m
(mean 2.1
m)
~20 に
676
Burns et al. (2017) Castlegate
Sandstone
& Neslen
Formation,
Utah, U.S.A.
に distal
splay
113 に 852
m; mean
399 m
(by 118 に
286 m;
mean 229
m)
0.2 に 1.6
m
(mean 0.8
m)
~71 に
4260
Lower
Beaufort
Group,
Karoo Basin,
South Africa
Permo-
Trias
sic
Semi-
arid
(seaso
nal)
Ephemeral deposits. Fine-
grained sandstone sheets
with a sharp or erosional
basal contact. Stage II and III
splay sets of Smith et al.
(1989) recognized in
proximal sequences.
> 600 m
to several
km along
strike
0.5 に 2 m < 1200
Stear (1983);
Smith (1987);
Jordaan (1990);
Smith (1993);
Gulliford et al.
(2014)
903
Table 3. Comparison of Beaufort Group crevasse splay architecture and geometry with 904
examples from modern and ancient fluvial systems. 905
83
Supplementary Information 906
Additional supplementary data may be found in the online version of this article: 907
Supplementary Figure 1. Overview maps and complete sedimentary log sections from 908
Verlatenkloof Pass (R354), near Sutherland (refer to Figure 1 and Supplementary Table 1 for 909
GPS coordinates). 910
Supplementary Figure 2. Overview maps and complete sedimentary log sections from 911
Müller Canyon, Gunsfontein, near Sutherland (refer to Figure 1 and Supplementary Table 1 912
for GPS coordinates). 913
Supplementary Table 1. Main locality names, abbreviations and coordinates relating to the 914
location of sedimentary logs within this study, listed alphabetically. 915
Supplementary Table 2. Database of proximal and distal splay thicknesses measured from 916
sedimentary logs at Verlatenkloof and Müller Canyon. Average thicknesses across these 154 917
proximal and distal splays is 110 cm and 45 cm, respectively. 918
Supplementary Spreadsheet 1: Excel spreadsheet comprising thickness, lithology, facies and 919
architectural element for beds measured at Verlatenkloof and Müller Canyon. Refer to 920
Figure 2 and Supplementary Table 1 for GPS coordinates, and Supplementary Figures 1 and 921
2 for Verlatenkloof and Müller Canyon sedimentary logs. This spreadsheet corresponds to 922
the sandstone to mudstone proportions and architectural elements pie charts presented in 923
Figure 3. 924